CN109285156A - The method that the solid end mill service life is used in offline prediction gear box of tractor processing - Google Patents

The method that the solid end mill service life is used in offline prediction gear box of tractor processing Download PDF

Info

Publication number
CN109285156A
CN109285156A CN201811368639.XA CN201811368639A CN109285156A CN 109285156 A CN109285156 A CN 109285156A CN 201811368639 A CN201811368639 A CN 201811368639A CN 109285156 A CN109285156 A CN 109285156A
Authority
CN
China
Prior art keywords
pixel
processing
cutter
cutting edge
eroded area
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Pending
Application number
CN201811368639.XA
Other languages
Chinese (zh)
Inventor
王建军
姚斌
宋世毅
马晓帆
陈彬强
游明琳
邓锦湖
孙志高
杨小勇
王裕喆
刘红霞
李晨硕
尹晨旭
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Xiamen University
First Tractor Co Ltd
Original Assignee
Xiamen University
First Tractor Co Ltd
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by Xiamen University, First Tractor Co Ltd filed Critical Xiamen University
Priority to CN201811368639.XA priority Critical patent/CN109285156A/en
Publication of CN109285156A publication Critical patent/CN109285156A/en
Pending legal-status Critical Current

Links

Classifications

    • GPHYSICS
    • G06COMPUTING; CALCULATING; COUNTING
    • G06TIMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T7/00Image analysis
    • G06T7/0002Inspection of images, e.g. flaw detection
    • G06T7/0004Industrial image inspection
    • GPHYSICS
    • G06COMPUTING; CALCULATING; COUNTING
    • G06TIMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T7/00Image analysis
    • G06T7/10Segmentation; Edge detection
    • G06T7/11Region-based segmentation
    • GPHYSICS
    • G06COMPUTING; CALCULATING; COUNTING
    • G06TIMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T7/00Image analysis
    • G06T7/10Segmentation; Edge detection
    • G06T7/187Segmentation; Edge detection involving region growing; involving region merging; involving connected component labelling
    • GPHYSICS
    • G06COMPUTING; CALCULATING; COUNTING
    • G06TIMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T2207/00Indexing scheme for image analysis or image enhancement
    • G06T2207/10Image acquisition modality
    • G06T2207/10004Still image; Photographic image
    • GPHYSICS
    • G06COMPUTING; CALCULATING; COUNTING
    • G06TIMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T2207/00Indexing scheme for image analysis or image enhancement
    • G06T2207/20Special algorithmic details
    • G06T2207/20024Filtering details
    • G06T2207/20032Median filtering
    • GPHYSICS
    • G06COMPUTING; CALCULATING; COUNTING
    • G06TIMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T2207/00Indexing scheme for image analysis or image enhancement
    • G06T2207/20Special algorithmic details
    • G06T2207/20081Training; Learning
    • GPHYSICS
    • G06COMPUTING; CALCULATING; COUNTING
    • G06TIMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T2207/00Indexing scheme for image analysis or image enhancement
    • G06T2207/30Subject of image; Context of image processing
    • G06T2207/30108Industrial image inspection
    • G06T2207/30164Workpiece; Machine component

Abstract

The method that the solid end mill service life is used in offline prediction gear box of tractor processing, is related to solid end mill.After solid end mill often processes the workpiece of setting quantity, with the picture of each cutting edge of cutting edge image acquisition device cutter;Eroded area is oriented using pixel as operating unit for the picture after acquisition, acquires eroded area pixel quantity;The relationship between tool sharpening number of packages amount and eroded area pixel reduction quantity is established, using the prediction realized from iteration Random Forest model to cutter life.After processing certain amount workpiece using solid end mill, stop processing, the chip on cutting edge is removed with bits blower, using cutting edge image acquisition device picture;Picture after processing acquisition positions eroded area, acquires eroded area pixel quantity using pixel as operating unit;With the prediction realized from iteration Random Forest model to cutter life.

Description

The method that the solid end mill service life is used in offline prediction gear box of tractor processing
Technical field
The present invention relates to solid end mills, more particularly to the solid end mill service life is used in offline prediction gear box of tractor processing Method.
Background technique
With the continuous propulsion that national intelligence manufacture " 2025 " plan, enterprise payes attention to product upgrading synergy further, and knife For tool as most important tool in machining, tool wear is to cause tool failure most heavy as tool failure main cause The influence factor wanted.
As research of the people to machine vision is more and more deep, ground based on the tool wear monitoring of machine vision in cutter Monitoring field is damaged using more and more.Machine vision observation tool wear is more intuitive, according to the observation than traditional worker workpiece The methods of surface, vibration and noise consider whether that tool changing is more accurate to judge the conventional method of tool wear.To existing Technical patent retrieval discovery, application No. is CN106312692A, denominations of invention are as follows: the tool wear inspection based on minimum circumscribed rectangle The Chinese invention patent application of survey method provide it is a kind of based on image recognition, using boundary rectangle acquire eroded area with it is undressed Cutter peak width ratio, judges tool wear degree.But there are the following problems for this method: (1) last eroded area square In shape fit procedure, because cutting edge profile shape is extremely complex, there are error of fitting, cause to judge tool wear degree accuracy Decline;(2) the only detection to current tool wear situation, and monitoring and prediction to cutter life is not implemented.
Summary of the invention
The purpose of the present invention is to provide the offline prediction gear box of tractor processing methods in solid end mill service life.
The present invention the following steps are included:
1) it is each with cutting edge image acquisition device cutter after solid end mill often processes the workpiece for setting quantity The picture of a cutting edge;Eroded area is oriented using pixel as operating unit for the picture after acquisition, acquires eroded area Pixel quantity;The relationship between tool sharpening number of packages amount and eroded area pixel reduction quantity is established, using gloomy at random from iteration Prediction of the woods model realization to cutter life.
In step 1), the application can to the specific method of the prediction of cutter life from the realization of iteration Random Forest model Are as follows: the service life state of solid end mill includes that tool wear early warning and tool wear repair state, uses S respectivelywAnd SrIt indicates, root Quantity is reduced according to tool in cutting sword pixel, is predicted in the piece count for reaching the tool sharpening that above two state of wear is predicted N is used respectivelypwAnd NprIt indicates, and the quantity N of current cutter actual processing workpiecerIt indicates, if Nr≤Npw, continue to use cutter and add Work;Npw< Nr≤Npr, prepare replacement cutter, guarantee processing efficiency;Nr> Npr, more renew cutter, the old cutter of reconditioning.
2) after processing certain amount workpiece using solid end mill, stop processing, on bits blower removal cutting edge Chip, using cutting edge image acquisition device picture;
In step 2), when the acquisition picture, it is accurate to position each cutting edge phase, cutter and phase when guarantee is taken pictures Machine relative position is identical.
3) picture after processing acquisition positions eroded area, acquires eroded area pixel number using pixel as operating unit Amount;
In step 3), the picture after the processing acquisition positions eroded area, acquires mill using pixel as operating unit The specific steps for damaging area pixel quantity can are as follows:
(1) picture is read in, pixel is subdivided with sub-pix processing method, improves image resolution ratio;
(2) to image median filter processing, select 9 × 9 templates to image median filter;
(3) profile is extracted in morphological images processing.To Image erosion processing, original image and corrosion image make poor image wheel It is wide;
(4) the connected domain operation removal small connected region of image.Eight connectivity operation, removes the small connected region in image;Eight Connection is defined as:
N8=N4∪(x+1,y+1),(x+1,y-1),(x-1,y+1),(x-1,y-1)
N in formula4For four connected region symbols, (x, y) is pixel P point coordinate.
(5) pixel coordinate is operated with geometric knowledge, determines cutting edge eroded area, before cutter is not used, existed first Determination two is without frayed edges curve K and L on cutter profile, it is assumed that the rectangular area of boundary curve K and L are respectively as follows:
X in formulaKmin,xKmax,yKmin,yKmaxRespectively indicate abscissa and ordinate in the rectangular area comprising boundary curve K Section maxima and minima, similarly hereinafter.
In view of the minor change of Image Acquisition phase, to improve fault-tolerance, (x, y) value will be in maximum value and minimum value Between, and there are enough surpluses.Identified two linearity region is defined as the region R and T, determines fit equation in the region R and T Are as follows: y=kx+b;
It is fitted with least square method and determines two linear equation lKAnd lL, simultaneous lKAnd lLEquation:
It is (x that two, which are acquired, without frayed edges straight-line intersectionKL,yKL);
A straight line region for having neither part nor lot in cutting is determined on cutting edge profile side:
To improve fault-tolerance, equally there are enough surpluses, utilize pixel coordinate fitting linear equation l in rectangular areaW Are as follows:
Y=kWx+bW
According to tool sharpening material, machined parameters, tool wear region geometry region is determined, determining principle generally should Region includes whole eroded areas of the knife face, and includes as few as possible non-eroded area, by lwLinear equation revolves counterclockwise Turn θ angle, obtains equation lX:
Y=tan (arctankw+θ)x+bX
Then by linear equation lK、lL、lXAnd lWAnd other regions of cutting edge constitute a closed eroded area Γ.
(6) according to different connected region areas, eroded area Γ is split, sum eroded area pixel quantity Np
(7) according to the above-mentioned specific steps for seeking eroded area pixel quantity, certain cutter Continuous maching N part workpiece is acquired, is cut It cuts sharpening damage area pixel and reduces data.
4) with the prediction realized from iteration Random Forest model to cutter life.It is different from traditional Random Forest model, Iterated conditional is added in model the present invention, reaches iteration convergence effect, improves the accuracy of prediction.
In step 4), it is described with from iteration Random Forest model realization can to the specific method of the prediction of cutter life Are as follows: it usesThe each cutting edge for indicating solid tool, uses κ123... indicate that cutting edge eroded area pixel subtracts Small number indicates the piece count of solid end mill processing with N, after solid end mill one workpiece of every processing, by pixel Point carries out geometric operation and obtains eroded area pixel reduction amount, it is assumed that after complete n-th of the workpiece of tool sharpening, is built according to above-mentioned parameter Mapping relations under Liru:
IfTo process the corresponding wear data of each cutting edge after n workpiece, if [1 23 ... n] be phase corresponding processing number of packages.
Experimental data data are divided into training sample data and forecast sample data, it is pre- using being returned from iteration random forest Method of determining and calculating is trained.It is different from traditional random forest regression forecasting, it changes being arranged from iteration random forest prediction algorithm For condition, data cleansing is carried out when meeting iterated conditional.
If four cutting edge eroded area pixel reduction amounts of cutter
What thus prediction solid end mill reached tool wear warning line can be with workpiece quantity for ni;And miTo reach knife Tool abrasion warning line actual processing number of packages amount, cutting edge eroded area pixel reduce quantity are as follows:
If ε is relaxation factor:
Indicate ζ14Indicate the iterated conditional and ζ of each cutting edge pixel reduction amountNFor the iteration of tool sharpening number of packages amount Condition, then:
ψ1=| pre-rea |
ψn=| ni-mi|
If ψ >=ε, shows to predict that the sample data accuracy of tool wear is poor, replace pre data with rea data, accordingly MiReplace ni, re -training sample data.In this way, being improved after introducing relaxation factor ε in traditional random forests algorithm The accuracy of training sample and prediction tool wear data.
The present invention is based on image minimum units, determine tool wear region with geometric knowledge, and tool wear region is straight It is expressed as pixel reduction amount with seeing;Based on above-mentioned data, a kind of cutter life prediction model is optimized, is realized to the cutter longevity The monitoring and prediction of life.
Detailed description of the invention
Fig. 1 is the flow chart of the embodiment of the present invention;
Fig. 2 is tool in cutting sword picture after sub-pix processing;
Fig. 3 is tool in cutting sword picture after median filter process;
Fig. 4 is image outline;
Fig. 5 is to remove image after small connected region;
Fig. 6 is the eroded area that boundary curve is constituted;
Fig. 7 is the eroded area split;
Fig. 8 is that pixel reduces quantity and workpiece processing tool Figure of the quantitative relationship.
Specific embodiment
Following embodiment will the present invention is further illustrated in conjunction with attached drawing.
1. processing gear box of tractor using solid end mill in certain model machining center, a workpiece image is often processed Acquisition device acquires the picture of each cutting edge.Acquisition device mainly includes locating piece and camera, and locating piece guarantees that milling cutter is every One cutting edge phase is correct, with the chip on bits blower removal cutting edge, acquires picture using camera, protects in collection process Cutter is identical as camera relative position when card is taken pictures, as shown in Figure 1.
2. the picture after processing acquisition positions eroded area, acquires eroded area pixel number using pixel as operating unit Amount.Specific steps are as follows:
1) picture is read in, pixel is subdivided with sub-pix processing method, improves image resolution ratio, as shown in Figure 2;
2) to image median filter processing.Select 9 × 9 templates to image median filter, image is as shown in Figure 3 after processing.
3) profile is extracted in morphological images processing.To Image erosion processing, original image obtains image as difference with corrosion image Profile, as shown in Figure 4;
4) the connected domain operation removal small connected region of image, as shown in Figure 5.Eight connectivity operation, removes the little Lian in image Logical region.Eight connectivity is defined as:
N8=N4∪(x+1,y+1),(x+1,y-1),(x-1,y+1),(x-1,y-1);
5) geometric knowledge is used, pixel coordinate is operated, determines cutting edge eroded area.Before cutter is not used, exist first Determination two is without frayed edges curve K and L on cutter profile, it is assumed that the rectangular area of boundary curve K and L are respectively as follows:
Two straight line regions are defined as R and T.Fit equation is determined in the region R and T are as follows:
Y=kx+b
Two linear equation l are fitted with least square methodKAnd lL.Simultaneous lKAnd lLEquation:
It is (1002,2697) that two, which are acquired, without frayed edges straight-line intersection.
A straight line for having neither part nor lot in cutting, rectangular area where straight line are determined on cutting edge profile side are as follows:
Utilize pixel coordinate fitting linear equation l in rectangular areaWAre as follows:
Y=0.7039x+26.7363
According to tool sharpening material, machined parameters, determine that tool wear geometric areas, determining principle are generally the region Whole eroded areas comprising the knife face, and as few as possible include non-eroded area.According to example tool wear region geometry Shape is determined lWLinear equation rotates 37 ° counterclockwise, obtains equation lX:
Y=-0.079x+1991.6922
By linear equation lK、lL、lXAnd lWAnd other boundary curves constitute closed eroded area Γ, such as Fig. 6 institute Show.
6) according to different connected region areas, eroded area Γ is split, as shown in Figure 7.Summation eroded area picture Prime number amount Np
7) workpiece of cutting edge Continuous maching 13 of cutter 3 being acquired according to the above method, eroded area pixel reduces data, As shown in Figure 8.
3. with the prediction realized from iteration Random Forest model to cutter life.It is different from traditional Random Forest model, Iterated conditional is added in model the invention, reaches iteration convergence effect, improves the accuracy of prediction.
According to the method described above, it usesThe each cutting edge for indicating solid tool, uses κ123... expression is cut It cuts sharpening damage area pixel and reduces quantity, the piece count of solid end mill processing, the every processing one of solid end mill are indicated with N After workpiece, eroded area pixel reduction amount is obtained by carrying out geometric operation to pixel, it is assumed that complete n-th of the workpiece of tool sharpening Afterwards, following mapping relations are established according to above-mentioned parameter:
IfTo process the corresponding wear data of each cutting edge after n workpiece, if [1 2 3 ... n] be phase corresponding processing number of packages.
Experimental data data are divided into training sample data and forecast sample data, it is pre- using being returned from iteration random forest Method of determining and calculating is trained.It is different from traditional random forest regression forecasting, it changes being arranged from iteration random forest prediction algorithm For condition, data cleansing is carried out when meeting iterated conditional.
If four cutting edge eroded area pixel reduction amounts of cutter
What thus prediction solid end mill reached tool wear warning line can be with workpiece quantity for ni.And miTo reach knife Tool abrasion warning line actual processing number of packages amount, cutting edge eroded area pixel reduce quantity are as follows:
If ε is relaxation factor:
Indicate ζ14Indicate the iterated conditional and ζ of each cutting edge pixel reduction amountNFor the iteration of tool sharpening number of packages amount Condition,
Then:
ψ1=| pre-rea |
ψn=| ni-mi|
If ψ >=ε, shows to predict that the sample data accuracy of tool wear is poor, replace pre data with rea data, accordingly MiReplace ni, re -training sample data.In this way, being improved after introducing relaxation factor ε in traditional random forests algorithm The accuracy of training sample and prediction tool wear data.
The prediction to cutter life is realized in the application from iteration Random Forest model.The service life state packet of solid end mill It includes tool wear early warning and tool wear repairs state, use S respectivelywAnd SrIt indicates.Quantity is reduced according to tool in cutting sword pixel, Prediction is respectively N in the piece count for reaching the tool sharpening that above two state of wear is predictedpwAnd Npr.And current cutter The quantity of actual processing workpiece is NrIf Nr≤Npw, continue to use tool sharpening;Npw< Nr≤Npr, prepare replacement cutter, guarantee Processing efficiency;Nr> Npr, more renew cutter, the old cutter of reconditioning.
The present invention proposes that a kind of pixel method that is based on predicts gear box of tractor processing integral vertical offline for tool wear The method in milling cutter service life, method include: with solid end mill process finished piece(s) after, it is each with cutting edge image acquisition device A cutting edge picture;Profile is extracted after the processing such as sub-pix processing, median filtering to each cutting edge picture, is with pixel Operating unit orients eroded area using geometric knowledge, according to the reduction quantity of eroded area pixel, judges the abrasion of cutter Situation;With being continuously increased for tool sharpening number of packages, the pixel quantity of eroded area can be reduced constantly, using gloomy at random from iteration The relationship of woods model foundation tool sharpening number of packages and eroded area pixel reduction amount realizes monitoring and prediction to cutter life.

Claims (5)

1. the method that the solid end mill service life is used in the processing of prediction gear box of tractor offline, it is characterised in that the following steps are included:
1) after solid end mill often processes the workpiece for setting quantity, with cutting edge image acquisition device cutter, each is cut Cut the picture of sword;Eroded area is oriented using pixel as operating unit for the picture after acquisition, acquires eroded area pixel Quantity;It establishes tool sharpening number of packages amount and eroded area pixel and reduces relationship between quantity, using from iteration random forest mould Type realizes the prediction to cutter life;
2) after processing certain amount workpiece using solid end mill, stop processing, with cutting on bits blower removal cutting edge Bits, using cutting edge image acquisition device picture;
3) picture after processing acquisition positions eroded area, acquires eroded area pixel quantity using pixel as operating unit;
4) with the prediction realized from iteration Random Forest model to cutter life, iterated conditional is added in model, iteration is reached Effect is restrained, the accuracy of prediction is improved.
2. the method that the solid end mill service life is used in offline prediction gear box of tractor processing, feature exist as described in claim 1 In in step 1), the prediction to cutter life is realized in the application from iteration Random Forest model method particularly includes: whole The service life state of slotting cutter includes that tool wear early warning and tool wear repair state, uses S respectivelywAnd SrIt indicates, is cut according to cutter It cuts sword pixel and reduces quantity, prediction uses N in the piece count for reaching the tool sharpening that above two state of wear is predicted respectivelypw And NprIt indicates, and the quantity N of current cutter actual processing workpiecerIt indicates, if Nr≤Npw, continue to use tool sharpening;Npw< Nr≤Npr, prepare replacement cutter, guarantee processing efficiency;Nr> Npr, more renew cutter, the old cutter of reconditioning.
3. the method that the solid end mill service life is used in offline prediction gear box of tractor processing, feature exist as described in claim 1 In in step 2), when the acquisition picture, position that each cutting edge phase is accurate, and cutter is opposite with camera when guarantee is taken pictures Position is identical.
4. the method that the solid end mill service life is used in offline prediction gear box of tractor processing, feature exist as described in claim 1 In in step 3), the picture after the processing acquisition positions eroded area, acquires eroded area using pixel as operating unit The specific steps of pixel quantity are as follows:
(1) picture is read in, pixel is subdivided with sub-pix processing method, improves image resolution ratio;
(2) to image median filter processing, select 9 × 9 templates to image median filter;
(3) profile is extracted in morphological images processing;To Image erosion processing, original image and corrosion image make poor image outline;
(4) the connected domain operation removal small connected region of image;Eight connectivity operation, removes the small connected region in image;Eight connectivity Is defined as:
N8=N4∪(x+1,y+1),(x+1,y-1),(x-1,y+1),(x-1,y-1)
N in formula4For four connected region symbols, (x, y) is pixel P point coordinate;
(5) pixel coordinate is operated with geometric knowledge, determines cutting edge eroded area, before cutter is not used, first in cutter Determination two is without frayed edges curve K and L on profile, it is assumed that the rectangular area of boundary curve K and L are respectively as follows:
X in formulaKmin,xKmax,yKmin,yKmaxRespectively indicate abscissa and ordinate section in the rectangular area comprising boundary curve K Maxima and minima, similarly hereinafter;
In view of the minor change of Image Acquisition phase, to improve fault-tolerance, (x, y) value will maximum value and minimum value it Between, and there are enough surpluses;Identified two linearity region is defined as the region R and T, determines fit equation in the region R and T Are as follows: y=kx+b;
It is fitted with least square method and determines two linear equation lKAnd lL, simultaneous lKAnd lLEquation:
It is (x that two, which are acquired, without frayed edges straight-line intersectionKL,yKL);
A straight line region for having neither part nor lot in cutting is determined on cutting edge profile side:
To improve fault-tolerance, equally there are enough surpluses, utilize pixel coordinate fitting linear equation l in rectangular areaWAre as follows:
Y=kWx+bW
According to tool sharpening material, machined parameters, tool wear region geometry region is determined, determining principle is that the region includes Whole eroded areas of knife face, and include non-eroded area, by lwLinear equation rotates θ angle counterclockwise, obtains equation lX:
Y=tan (arctankw+θ)x+bX
Then by linear equation lK、lL、lXAnd lWAnd other regions of cutting edge constitute a closed eroded area Γ;
(6) according to different connected region areas, eroded area Γ is split, sum eroded area pixel quantity Np
(7) according to the above-mentioned specific steps for seeking eroded area pixel quantity, certain cutter Continuous maching N part workpiece, cutting edge are acquired Eroded area pixel reduces data.
5. the method that the solid end mill service life is used in offline prediction gear box of tractor processing, feature exist as described in claim 1 It is described with the prediction realized from iteration Random Forest model to cutter life in step 4) method particularly includes:
WithThe each cutting edge for indicating solid tool, uses κ123... indicate that cutting edge eroded area pixel subtracts Small number indicates the piece count of solid end mill processing with N, after solid end mill one workpiece of every processing, by pixel Point carries out geometric operation and obtains eroded area pixel reduction amount, it is assumed that after complete n-th of the workpiece of tool sharpening, is built according to above-mentioned parameter Mapping relations under Liru:
IfTo process the corresponding wear data of each cutting edge after n workpiece, if [1 23 ... N] be phase corresponding processing number of packages;
Experimental data data are divided into training sample data and forecast sample data, are calculated using from iteration random forest regression forecasting Method is trained;Iterated conditional is being set from iteration random forest prediction algorithm, it is clear that data are carried out when meeting iterated conditional It washes, if four cutting edge eroded area pixel reduction amounts of cutter
Thus it is n that prediction solid end mill, which reaches the workpiece quantity of tool wear warning line,i;And miIt is alert to reach tool wear Line actual processing number of packages amount is guarded against, cutting edge eroded area pixel reduces quantity are as follows:
If ε is relaxation factor:
Indicate ζ14Indicate the iterated conditional and ζ of each cutting edge pixel reduction amountNFor the iterated conditional of tool sharpening number of packages amount, Then:
ψ1=| pre-rea |
ψn=| ni-mi|
If ψ >=ε, shows to predict that the sample data accuracy of tool wear is poor, replace pre data, corresponding m with rea datai Replace ni, re -training sample data.
CN201811368639.XA 2018-11-16 2018-11-16 The method that the solid end mill service life is used in offline prediction gear box of tractor processing Pending CN109285156A (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN201811368639.XA CN109285156A (en) 2018-11-16 2018-11-16 The method that the solid end mill service life is used in offline prediction gear box of tractor processing

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN201811368639.XA CN109285156A (en) 2018-11-16 2018-11-16 The method that the solid end mill service life is used in offline prediction gear box of tractor processing

Publications (1)

Publication Number Publication Date
CN109285156A true CN109285156A (en) 2019-01-29

Family

ID=65176378

Family Applications (1)

Application Number Title Priority Date Filing Date
CN201811368639.XA Pending CN109285156A (en) 2018-11-16 2018-11-16 The method that the solid end mill service life is used in offline prediction gear box of tractor processing

Country Status (1)

Country Link
CN (1) CN109285156A (en)

Cited By (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN109894922A (en) * 2019-03-07 2019-06-18 中南大学 It is a kind of it is accurate broached in chip out of phase thickness method
CN111519506A (en) * 2020-04-28 2020-08-11 三一汽车制造有限公司 Milling cutter wear prompting method and milling machine

Citations (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN102501140A (en) * 2011-11-22 2012-06-20 南京航空航天大学 Method for positioning and monitoring wear of ball nose end mill cutter
KR101621109B1 (en) * 2010-11-26 2016-05-13 토포인트 테크날러지 코포레이션 리미티드 Testing method for machining cutter
CN106944877A (en) * 2016-01-06 2017-07-14 哈尔滨理工大学 Tool wear intelligent wireless detection means based on acoustic emission signal
CN107717631A (en) * 2017-10-31 2018-02-23 中国科学院合肥物质科学研究院 A kind of HSC Milling Tools wear automatic monitoring method
CN108062072A (en) * 2017-11-23 2018-05-22 哈尔滨理工大学 A kind of dual camera image collecting device and the flat-bottom milling cutter of image mosaic abrasion online test method
CN108536938A (en) * 2018-03-29 2018-09-14 上海交通大学 A kind of machine tool life prediction system and prediction technique

Patent Citations (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
KR101621109B1 (en) * 2010-11-26 2016-05-13 토포인트 테크날러지 코포레이션 리미티드 Testing method for machining cutter
CN102501140A (en) * 2011-11-22 2012-06-20 南京航空航天大学 Method for positioning and monitoring wear of ball nose end mill cutter
CN106944877A (en) * 2016-01-06 2017-07-14 哈尔滨理工大学 Tool wear intelligent wireless detection means based on acoustic emission signal
CN107717631A (en) * 2017-10-31 2018-02-23 中国科学院合肥物质科学研究院 A kind of HSC Milling Tools wear automatic monitoring method
CN108062072A (en) * 2017-11-23 2018-05-22 哈尔滨理工大学 A kind of dual camera image collecting device and the flat-bottom milling cutter of image mosaic abrasion online test method
CN108536938A (en) * 2018-03-29 2018-09-14 上海交通大学 A kind of machine tool life prediction system and prediction technique

Non-Patent Citations (2)

* Cited by examiner, † Cited by third party
Title
秦国华,易鑫,李怡冉,谢文斌: "刀具磨损的自动检测及检测系统", 《光学精密工程》 *
赵帅,黄亦翔,王浩任,刘成良,刘晓,梁鑫光: "基于随机森林与主成分分析的刀具磨损评估", 《机械工程学报》 *

Cited By (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN109894922A (en) * 2019-03-07 2019-06-18 中南大学 It is a kind of it is accurate broached in chip out of phase thickness method
CN111519506A (en) * 2020-04-28 2020-08-11 三一汽车制造有限公司 Milling cutter wear prompting method and milling machine

Similar Documents

Publication Publication Date Title
CN109285156A (en) The method that the solid end mill service life is used in offline prediction gear box of tractor processing
Kerr et al. Assessment and visualisation of machine tool wear using computer vision
CN102353680B (en) Method for evaluating surface detects of small-sized workpieces and flow for detecting unqualified workpieces
CN106872476A (en) A kind of casting class workpiece surface quality detection method and system based on line-structured light
CN107877262B (en) A kind of numerical control machine tool wear monitoring method based on deep learning
Chen et al. Automated inspection of engineering ceramic grinding surface damage based on image recognition
Datta et al. Progressive cutting tool wear detection from machined surface images using Voronoi tessellation method
WO2014091884A1 (en) Lens processing system, tool change time detection method, and eyeglass lens manufacturing method
CN109060836B (en) Machine vision-based high-pressure oil pipe joint external thread detection method
Wu et al. Watershed-based morphological separation of wear debris chains for on-line ferrograph analysis
CN107742289A (en) One kind is based on machine vision revolving body workpieces detection method
CN104751474A (en) Cascade quick image defect segmentation method
Lutz et al. Evaluation of deep learning for semantic image segmentation in tool condition monitoring
CN108490880B (en) Method for monitoring wear state of cutting tool of numerical control machine tool in real time
Tiwari et al. Tool wear prediction in end milling of Ti-6Al-4V through Kalman filter based fusion of texture features and cutting forces
CN109101976A (en) A kind of detection method of arc extinguishing grid pieces surface defect
Oo et al. Tool wear monitoring system in belt grinding based on image-processing techniques
Liu et al. Real-time cutting tool state recognition approach based on machining features in NC machining process of complex structural parts
Chen et al. Notice of Retraction: Application of optical inspection and metrology in quality control for aircraft components
CN110587377B (en) Method for monitoring defects of milling cutter on line
Mook et al. Measurement of nose radius wear in turning tools from a single 2D image using machine vision
Kassim et al. Machine tool condition monitoring using workpiece surface texture analysis
CN105405137A (en) Shaft quality detection method based on machine vision
Liang et al. Vision-based automatic tool wear monitoring system
Kassim et al. Texture analysis using fractals for tool wear monitoring

Legal Events

Date Code Title Description
PB01 Publication
PB01 Publication
SE01 Entry into force of request for substantive examination
SE01 Entry into force of request for substantive examination